Real-time data Helps Yielding Traffic Handling Models

One of the basic tasks of every motorway-network Traffic Control Centre (TCC) is to provide accurate descriptions of the current traffic conditions inside the network. In order for this task to be accomplished, a number of sensors such as magnetic loop detectors, video cameras, floating cars, etc., provide data to the TCC, which are used for traffic surveillance and control purposes. Appropriate algorithms for state estimation/prediction, travel time estimation/prediction, queue-tail tracking, incident detection, and traffic control are subsequently employed for comprehensive surveillance and control of the network traffic flow.

RHYTHM aims at developing new algorithms, or modifying already existing ones, for traffic surveillance and control that use video-based data and rigorously compare them with algorithms that use data obtained by use of loop detectors. Video sensors have proved advantages over traditional loop detectors for various specific applications beyond the mere measurement of traffic variables, e.g. for incident detection. For the further deployment of traffic sensors, however, it is important to test their efficiency for further specific surveillance and control tasks of practical importance, and to compare with loop-based efficiency. As a matter of fact, the use and full exploitation of video-based information calls for suitable modifications or even for complete re-design of available loop-based surveillance or control algorithms along with field application, demonstration, and comparative evaluation.